Here’s a sample of some of the things we’re working on at UXLabs this year, neatly packaged into Masters level ‘internships’. I use quotes there as although it’s a convenient term used by many of my academic colleagues, these opportunities are (a) unpaid and (b) remote (i.e. hosted by your own institution). So maybe ‘co-supervised MSc projects initiated by a commercial partner’ is more accurate term… Anyway, what we offer is support, expertise, co-supervision and access to real world data/challenges. If you are interested in working with us on the challenges below, get in touch.
Interactive Query Expansion
According to the IDC whitepaper, The High Cost of Not Finding Information, knowledge workers spend 2.5 hours per day searching for information. Assuming they either find what they are looking for eventually or stop and end up making a non-optimal decision, there is a high cost to both outcomes. To address this problem, UXLabs is investigating alternatives to traditional query formulation. We are developing a novel framework which allows users to express complex information needs via a simple but powerful visual syntax. A key part of this is the task of keyword selection: for a given concept, how can we help users identify an optimal set of search keywords and phrases? The aim of this project is to develop an interactive query generation engine which can expand an initial expression to include an optimal set of related terms and synonyms, going beyond the limitations of traditional auto-completion techniques.
Developing a robust model of query reformulation
Query reformulation is a common part of users’ information retrieval behaviour, whereby search queries are adapted until the user fulfils their information need or abandons their search. Previous studies have shown that users display distinct patterns of query reformulation, and that these patterns can be used to define behavioural habits specific to individual users. However, previous approaches have generally been restricted to using patterns based on simple lexical transformations, i.e. adding or removing words & characters. In this project, we seek to extend that work to address more conceptual relationships such as those due to semantics or other types of higher-level association. This could involve the use of knowledge based resources such as WordNet and/or machine readable dictionaries & thesauri. The outcome would be a more robust model of query reformulation behaviour that more accurately reflects the search habits and preferences of individual users.
A language for search and discovery
There has been much work on describing and modelling human information seeking behaviour. In one particular model, information behaviour is represented as a set of search modes that users employ to satisfy their information needs, such as ‘locate’, verify’, ‘monitor’, and so on. An interesting property of this framework is that modes do not occur randomly, but instead are seen to form distinct patterns or chains of information seeking behaviour. The goal of this project is to analyse a collection of real-world information needs and identify and visualize the patterns within them. In so doing we would hope to explore how the modes can be used as an elementary ‘language’ for describing information behaviour. We also hope to extend the framework and explore its ability to accommodate new domains and scenarios.
- Prostitutes Appeal to Pope: Text Analytics applied to Search
- The role of Natural Language Processing in Information Retrieval
- Extracting sentiment from healthcare survey data
- Mining search logs for usage patterns (part 2)
- Sentiment analysis: a comparison of four tools